The most appropriate next step for this woman with intermittent heavy vaginal bleeding is referral for endometrial biopsy. Her bleeding is characteristic of an anovulatory pattern based on the unpredictable occurrence of bleeding of variable flow and duration. In women with prolonged anovulation, there is loss of normal hormonal flux with exposure to unopposed estrogen without the normal endometrial protective effect of progesterone. This increases the risk for endometrial hyperplasia and endometrial malignancy. Additional risk factors for endometrial cancer in premenopausal women include obesity, nulliparity, age 35 years or older, diabetes mellitus, family history of colon cancer, infertility, and treatment with tamoxifen. In women younger than 35 years of age with anovulatory bleeding and no other risk factors for endometrial cancer, hormonal therapy for anovulation is appropriate. However, in women younger than 35 years of age with risk factors, or any patient with anovulatory bleeding 35 years of age or older, endometrial biopsy should be performed to exclude significant endometrial pathology. Because this patient is 40 years of age and has an additional risk factor for endometrial cancer (nulliparity), endometrial biopsy is the most appropriate next step in management.

A follicle-stimulating hormone (FSH) level can be used to evaluate ovarian dysfunction, but levels vary depending on menstrual phase, age, medications, and hormonal disorders. FSH level can be used to confirm menopausal status if a woman has been amenorrheic for longer than 12 months. However, a single FSH level can be misleading during perimenopause because levels may vary. In this patient, who is premenopausal, measurement of the FSH level is of limited use.

In a woman of reproductive age with abnormal vaginal bleeding, the first test to be performed is always a urine pregnancy test, which is a sensitive qualitative measure of β-human chorionic gonadotropin level and was already obtained in this patient. There is no further gain from obtaining a serum β-human chorionic gonadotropin level, which is a quantitative measure but adds no further value in this diagnostic evaluation.

Transvaginal ultrasonography may have a role in evaluation of postmenopausal bleeding to assess for endometrial thickness, but it is not useful in assessing bleeding in premenopausal women due to significant variations in endometrial thickness caused by hormonal fluctuation. It is therefore not used routinely for evaluation of premenopausal bleeding unless a structural uterine abnormality that may be contributing to bleeding is suspected.